Background/Aims: Spontaneous bacterial peritonitis(SBP) is a serious and frequently fata1 compli- cation in cirrhotics, therefore earlier detection and earlier institution of appropriate treatment is crucial for good result. However, the conventional method of ascitic tluid culture has a low sensitivity in detecting causative organism of SBP. Methods: We have compared the sensitivity of the conventional method to the bedside inoculation of ascites into blood culture media. Two kinds of ascites culture methods were compared in cirrhotic patients with spontaneous bacterial peritoni- tis: 1) conventional(on blood agar, MacConkey agar, thioglycolate broth, and phenylethanol blood agar) and 2) blood culture media method(inoculation of each S mL of ascites into one 30-mL tryptic soy broth bottle and one30-mL thioglycolate broth bottle at the patients bedside. Results: In a 55-month period, 67 episodes of SBP in 61 cirrhotic patients were examined using simultane- ous both culture methods. The conventional method grew bacteria in 22 episodes(32.8%), whereas the blood culture media method grew in 55 episodes(82.1%), which showed a significantly higher sensitivity(p0.005). The organisms most frequently isolated in ascitic fluids in this study were aerobic gram-negative bacteria(88.37o), among them E. coli was the most most common pathogen comprising 55.0% of total isolates. The conventional method grew five species of organism whereas the b]ood culture media did eleven species. Most of E. coli were susceptible to amino- glycosides, third generation cephalosporins, aztreonam, and ofloxacin, but resistant to ampicil]in. Conclusions: It is suggested that the inoculation of ascitic fluid to a blood culture media is more sensitive than the conventional method and shou]d be routinely used for ascitic fluid culture in cirrhotic patients. (Korean J Gastroenterol 1995;27: 659 - 672)